Table 3 Results of stepwise, linear regression models for changes in VAAE, IOD-VA for the anisometropic group (n = 97 cases). Stereoacuity data at baseline were not gathered for the anisometropic amblyopes. The importance (I) of a predictor is the residual sum of squares with the predictor removed from the model, normalized so that the importance values sum to 1. The percentage is the model’s accuracy, defined as the adjusted R². C: coefficient; Significant p-value (*, p < 0.05). Cells for model terms not contributing significantly to the model’s predictive power are empty. The factors missing from the table, do not contribute to the model. The numbers between parentheses indicate the order of importance in the model for each statistically significant variable

From: Factors affecting the benefit of glasses alone in treating childhood amblyopia: an analysis of PEDIG data